Domestic violence is a well recognized public health problem, and court orders provide women with the opportunity to use a legal intervention to prevent future contact with an abusive partner. The effectiveness of court orders is unclear, as prior studies have been descriptive case series or of small size. The proposed cohort study will evaluate the effectiveness of court orders in preventing subsequenting violence and injury in two novel ways: (1) a comparison group of abused women who did not seek legal protection will be included and (2) medical data as well as criminal justice data and survey information will be used to evaluate outcomes. The cohort will consist of female Seattle residents with police or court domestic violence contact during 1995-1997. The subgroups will be victims who obtained a protection or no contact order, compared to victims who did not obtain either order. The outcomes of interest (recurrent violence, injury, medical care, hospitalization, and death) will be ascertained from police and court records, emergency medical service data, emergency room and hospitalization records, and death certificates. Additionally, a subset of women entering the cohort during 1997 will be surveyed on entry about socio-demographic characteristics, relationship and abuse history, details of the index violent event, and social support system. For a random sample of the 1997 women, 3-month and 12-month telephone follow-up surveys will obtain information on the frequency and type of subsequent abuse, type and severity of injury, frequency and types of violations, and enforcement of violations, self-perception of well-being, support and advocacy services. Logistic regression will be used to evaluate the effectiveness of and the characteristics of court orders and other services in preventing subsequent contact, violence, and injury, while controlling for relevant confounding variables such as abuse history. The findings will be useful to police, courts, and health care providers in making necessary changes to improve the effectiveness of such interventions.